Letter: What is the Marginal Cost of Using Robot Assistance or Navigation for Transforaminal Lumbar Interbody Fusion? A Time-Driven Activity-Based Cost Analysis

摘要

To the Editor: We read with great interest the recent article by Sarikonda et al1 comparing the cost of transforaminal lumbar interbody fusion using freehand (n = 233), intraoperative navigation (n = 59), or robotic-assisted techniques (n = 20). The authors found that robotic surgeries had significantly higher costs than nonrobotic cases, attributed largely to increased supply costs amounting to nearly $8000-$9000 compared with intraoperative navigation and freehand surgeries. On multivariable regression, personnel costs did not differ significantly between the cases. We commend the authors for their important work and use of time-driven activity-based costing analysis to shed light on an often-obscure topic. Cost-effectiveness is an important consideration for hospitals, physicians, and patients deciding on the optimal technique. We have no doubt as to the rigor of the authors' data and statistical methodology but would like to highlight our own institutional experience with robotic surgery and offer several caveats to the study. The robotic platform requires an upfront investment of capital, but additional expenses incurred for each robotic case typically consist only of sterile drapes for the robot and fiducials (around $200). At our institution, we are not aware of any differences in the costs of spinal instrumentation based on the presence or absence of an intraoperative robotic platform. If different manufacturers are used between robotic and nonrobotic cases, prices are expected to differ, although enterprise pricing usually keeps various implant costs similar at a given institution. Institutions that significantly incorporate robots into their surgical armamentarium or use universal vendors for supplies may negotiate better prices for instrumentation than those that do not routinely use robots or rely on many vendors; however, this price reflects contractual differences and is not an inherent feature of robotic surgery. Some of the findings in the study may reflect the limited sample size of the robotic cohort and the authors' early experience with robotic surgery, when the learning curve is at its steepest. They found that expenses for robotic cases included a higher share of consumables, such as sterile towels and gauze, compared with nonrobotic cases. We see no reason for this discrepancy, particularly once the initial learning curve is surmounted. Indeed, previous studies have shown significant reductions in blood loss using robots, which should translate to fewer consumables.2 The authors also found increased operative time with robotic surgery compared with other modalities. However, we have shown progressive reductions in operative time as surgeons gain expertise with robotic surgery, until mastery is reached after around 67 cases.3 Other studies have reported no differences in time in the operating room.2,4 It would be interesting to apply the time-driven activity-based costing analysis to surgeons who have accumulated a large cohort of robotic cases and can be considered to achieved mastery. Robotic platforms offer other advantages that reduce costs in the long run but are not immediately reflected in the operative charges. Systematic reviews have shown that patients spend less time hospitalized after robotic surgery compared with freehand pedicle screw insertion.2,5 Each day of hospitalization adds to the patients' total costs. Robots also have the important benefit of reducing radiation exposure compared with navigation and freehand techniques.4,6 Moreover, overall rates of complications are reduced with robotic surgery, including infection, misplaced hardware, hardware failure, surgical revision, and proximal-facet joint violations.6 The level of health care spending on surgical complications avoided due to robotic surgery can be difficult to calculate but should be factored into decision-making about the cost-effectiveness of a robotic platform. Finally, we reiterate that a principal reason we prefer robot